The effort to understand and influence brand buying decisions is not new. In fact, it has been a primary focus of brand managers, researchers and marketers for decades and represents a substantial chunk of yearly budgets. Billions of dollars per year, and trillions in total, have been spent trying to predict demand and preferences, drive trial, inspire loyalty and encourage shoppers to switch from the competition. Big investments typically indicate high stakes, and this is no exception. Studies have found that 90% of the top 100 brands lost category share in 2014/15 as competition from private labels, product proliferation and more diverse consumer preferences continue to increase the pressure on big brands.
The current state of the art relies heavily on two key disciplines and related data sources—consumer behavior and after-the-fact performance tracking. In broad strokes, the field of consumer behavior provides an understanding of consumer preferences and attitudes toward brands and products and helps to define consumer needs and wants. The combination of brand affinity assessment and needs analysis drives decisions in a wide range of areas, including brand marketing, new product development, packaging and pricing.
The second key source of market feedback is grounded in sales data and consumer-reported consumption data and is used as a yardstick for measuring brand performance and the impact of the huge budgets spent to move the sales needle. These data sources provide a coarse feedback option for tracking changes in volume and predicting demand for a particular brand.
Traditional methods have been able to capture the two endpoints comprising what someone might want or need and a sample of what shoppers actually purchased. These methods, however, provide little to no insight as to what happens in between. This creates a need, therefore, to determine in-store shopping and buying behavior by various shopper segments. This need is particularly felt with regards to determining whether the purchase decisions were made before the shopper encountered the product at the shelf (pre-shelf) or while the shopper was present at the shelf (at-shelf).
Information regarding the decision process can be obtained in a number of ways, such as via surveys or shopper interviews. These methods require participation from the shopper, however, introducing the possibility of bias into the results. Also, practical limitations dictate that there is a limit on sample size. Therefore, there is also a need to automatically capture decision data for a large number of shoppers without voluntary participation by those shoppers.